Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
In their comment, Taylor, Kreisle and Zimmerman use gasoline price data taken from fleet card transactions at selected gasoline stations to re-examine a subset of results presented in Hastings (2004). Bringing new data to re-examine the question is a helpful contribution. Both data sets have limitations, potentially causing differences in the estimated effect. I worked with the authors to explore and understand the differences in the data sets and how they impact the estimates in both analyses, and conclude that the effects sizes are likely smaller in areas of overlap between the two data sets.